NVIDIA Software Announces Software Engineering Internship for 2026: Work on CUDA Core Libraries — Germany & France (Remote Possible)

NVIDIA, the pioneer of GPU computing and the engine behind the modern AI revolution, has opened applications for its Software Engineering Internship - CUDA Core Libraries for Summer 2026. This is a world-class opportunity based in Munich, Germany, designed for students who want to build the foundational software that powers high-performance computing (HPC) and AI.

This full-time internship is open to BS, MS, and PhD students. If you are passionate about systems-level development, C++, and Python, this role places you at the heart of NVIDIA's accelerated computing platform. You will contribute to the libraries that enable developers worldwide to write fast, scalable GPU-accelerated software.

Internship Overview & Key Details

Role Responsibilities: What Will You Do?

As an intern on the CUDA Core Libraries team, you will not just be writing code; you will be defining how code is written for GPUs. You will work on critical projects like CCCL (Thrust, CUB, libcudacxx), cuda-python, and numba-cuda. Your core duties will include:

• Library Design: You will contribute to the design and implementation of CUDA Core Libraries in C++ and Python, focusing on parallel algorithms.
• Optimization: You will design and optimize GPU algorithms and APIs, handling low-level performance tuning involving memory, parallelism, and synchronization.
• Developer Experience: You will improve the ecosystem by working on tests, benchmarks, Continuous Integration (CI), and documentation.
• Collaboration: You will participate in design reviews and code reviews with experienced CUDA engineers in an open-source-style workflow.

Eligibility and Mandatory Requirements

NVIDIA is looking for outstanding engineers who understand systems-level development. To qualify, you must meet the following criteria:

Mandatory Qualifications

• Education: You must be currently pursuing a BS, MS, or PhD in Computer Science, Computer Engineering, or a related field.
• Programming Skills: Strong proficiency in C++, Python, or both is required. You should have an interest in performance, memory, and concurrency.
• Systems Knowledge: Familiarity with modern C++ (templates, standard library) and/or Python library packaging.
• Parallel Programming: Experience with parallel or heterogeneous programming (e.g., CUDA, OpenMP) through coursework or projects is essential.

How to Stand Out (Preferred Skills)

• Knowledge of CPU/GPU architecture and algorithmic performance.
• Hands-on experience with Thrust, CUB, or libcudacxx.
• Familiarity with compiler infrastructure like LLVM or MLIR.
• Experience debugging large, multi-language codebases (CMake, GitHub Actions).

Why Apply?

Interning at NVIDIA means working with the smartest people in the industry on problems that matter. You will help build the tools that drive deep learning, scientific computing, and data analytics. It is a challenging, autonomous environment where your code could accelerate the next big breakthrough in AI.

How to Apply

Applications for the Summer 2026 program are now open. Interested candidates should submit their resume (in English) via the NVIDIA careers portal. Be sure to highlight your systems programming projects and any experience with GPU acceleration.

Official Application Link: Click Here to Apply for NVIDIA CUDA Internship